The emerging subfield of “security demographics” is interested in how demographic trends, such as youth bulges, high or low fertility rates, and sex ratios affect the security and stability of nation-states and regions. In our research, Andrea Den Boer and I have attempted to show that abnormally high sex ratios – situations where there are significantly more men than women – have been a security concern in the past and may affect security and stability in the future.

In China, for example, there are approximately 120 baby boys born for every 100 baby girls, and depending on which scholar you consult, there are already between 30 and 50 million more men than women, most of them in the younger generations. These abnormal sex ratios are primarily due to cultural and policy-derived incentives that can lead to sex-selective abortion of female fetuses.

In our book, Bare Branches: The Security Implications of Asia’s Surplus Male Population, Den Boer and I argue higher proportions of involuntarily unmarried men contribute to higher levels of violent crime, substance abuse, and collective violence because such men experience physiological, social, and status discrepancies that predispose them to risky, anti-social behavior. Government responses to this internal threat may have ramifications for levels of nationalism within the society, as well as the calculus of deterrence between societies. In sum, creating a high sex ratio society is not a move that enhances stability and security.

Since Den Boer and I have done a great deal of work demonstrating this link between high sex ratios and violence, when a colleague recently sent along a new paper published in Trends in Ecology and Evolution arguing the opposite – that societies with abnormal sex ratios favoring males do not exhibit higher levels of violence – I was intrigued. Indeed our work is the first citation in this article, and others referenced include well-known researchers who have made the same claim, such as Richard Wrangham, Martin Daly, Margo Wilson, Jean Dreze, and Satoshi Kanazawa.

What Matters More: Marriage or Mating Conflict?

The authors, University of California Davis anthropology graduate students Ryan Schacht and Kristin Liv Rauch and their mentor Monique Borgerhoff Mulder, examine three reasons why male-oriented sex ratios might lead to violence:

• A: Men commit more violent acts, so more men means more violence;
• B: The higher the proportion of specifically unmarried men, the higher the violent crime rate; or
• C: Increased mating competition (fewer eligible women than men) produces more male-on-male mating conflict.

“A” is not very interesting, so the authors leave it alone, and I would agree with that choice. “B” is the theory that Den Boer and I have worked with over the years, but the authors do not engage it. “C” is the theory the authors discuss most thoroughly.

Engaging theory C, Schacht et al. suggest that in high sex ratio human societies one should see pressure for monogyny (males mate with one female) and greater paternal investment in children as forms of “paternity protection,” therefore potentially reducing conflict by males over mates. I think we do see more pressure for monogyny in human societies with higher sex ratios and paternal investment; however, these are not logically inconsistent with theory B, that having a much higher proportion of unmarried males in society might produce higher levels of violence. Neither is it logically inconsistent with theory C if “paternity protection” also necessitates the use of violence.

Next, the authors look at adult sex ratio and “opportunity for sexual selection,” which is – if you search hard enough to find it – operationalized as higher variability in the number of offspring between individual men in a society. They find that adult sex ratio and “opportunity for sexual selection” are related negatively. However, no significance levels are provided and the confidence intervals are so large there may be no relationship at all. Again, this move does not engage theory B, and arguably only tangentially touches on theory C, which is about male-on-male mate competition specifically and not differences between men with regard to the number of children they have. These concepts, though possibly related, are not the same. (See addendum for additional technical comments on this point.)

A Non-Linear Relationship

The authors survey the empirical and quasi-empirical literature examining sex ratios and violence, providing a large table of comparisons. While they note that how sex ratio is defined varies across the studies, the actual measures are not provided in the table. For each study, the authors assess whether the study finds a positive or negative relationship between sex ratios and violence.

The authors conclude there is “mixed and contradictory” evidence for the proposition that high sex ratios are associated with higher levels of violence. What could account for such mixed results? It is well known that abnormal sex ratios in either direction (too high or too low) result in elevated violence. In other words, the relationship is U-shaped (see Figure 1).

However, the authors look for a linear relationship – that is, they ask whether high sex ratios are associated with higher levels of violence, and low sex ratios with lower levels of violence.

You will not find evidence for a linear relationship if the relationship is non-linear. Almost half the studies in the authors’ table examine sub-national level data from the United States, which contains significant pockets of very low sex ratios (more women than men) coupled with very high rates of violence. These areas – typically poor inner-city neighborhoods – have been intensively studied. Given what we know about the U-shaped relationship between sex ratio and violence, we now see why they find “mixed and contradictory evidence”: you cannot use a linear model to capture a non-linear relationship.

Street Crime vs. Antler Locks

To decide if high sex ratios are a risk factor for heightened violence – an important question for many countries – one must compare high sex ratio societies to normal sex ratio societies, not to low sex ratio societies. There are numerous rigorous empirical studies showing that areas in India and China with higher sex ratios have significantly higher violent crime rates than areas with more normal sex ratios.

The authors suggest that one key flaw in the proposition that high sex ratios lead to high levels of violence is that not all male aggression stems from male-on-male mate competition, which is pertinent specifically to theory C. While completely true, this is also beside the point if you are working with theory B. A random mugging is not like one male deer locking antlers with another over a doe (but it may well have something to do with the mugger’s unmarried state).

While I am ecstatic to see more work on the link between sex ratios and violence, and while the authors’ work is very creative, their sweeping conclusion that high sex ratio societies are not more violent is not supported by their reasoning or their evidence. A lot more work remains to be done to determine under what conditions this relationship holds in the real world; but we need to design research that connects more fully with the research question.